Raw materials qualification within a workflow: FT-NIR analysis using the Antaris II Analyzer

Importance of the topic

Fourier-transform near-infrared (FT-NIR) analysis at goods-in is a high-impact element of Process Analytical Technology (PAT). Rapid, non-destructive raw material identification and basic physical characterization at the factory entry point protect downstream product integrity, reduce quarantine time, and lower the burden on conventional laboratory testing. Implementing FT-NIR with an automated workflow enables timely release decisions, deeper traceability, and tighter integration with manufacturing control systems and inventory management.

Objectives and overview of the application note

This application note demonstrates a model workflow that integrates the Thermo Scientific Antaris II FT-NIR Analyzer with RESULT software to qualify incoming raw materials. The objectives were to: identify chemical composition non-destructively (including through packaging), discriminate material classes, evaluate particle-size-related spectral effects, and automate decision making and reporting to plant control systems. The study illustrates chemometric approaches and workflow events that enable unattended, goods-in testing at the loading dock or warehouse.

Methodology and workflow

• Workflow architecture: RESULT software organizes analysis into programmable events (Request, Measure, Perform If, archive/report). Input can be manual, barcode, text files, or process/control systems.
• Sampling approach: SabIR diffuse-reflectance fiber probe attached to the Antaris II enables spectral collection through drum liners, capturing both liner and sample signatures without unpacking.
• Chemometric strategy: Parallel comparison of unknown spectra against multiple libraries ("Classify Multiple"). Qualitative algorithms used include Similarity Match, QC Compare, Distance Match, and Discriminant Analysis. TQ Analyst chemometrics (SIMCA, PCA, Mahalanobis distance) provide classification and outlier detection.
• Preprocessing: Multiplicative Scatter Correction (MSC), second derivative transformation, and Norris smoothing (segment length 9, gap 3) were applied as needed to suppress scattering and noise. Specific polyethylene liner absorptions (~4200–4350 cm^-1 and ~5650–5800 cm^-1) were excluded from analysis.
• Decision logic: If a match in the library is found, the system issues a PASS, archives data, and reports via OPC to DCS and to LIMS/inventory. Non-matching materials generate FAIL, route the sample to quarantine, and notify the supplier.

Used instrumentation

• Antaris II FT-NIR Analyzer (Thermo Scientific)
• SabIR diffuse reflectance fiber probe for non-destructive sampling through packaging
• RESULT software (Thermo Scientific) for workflow automation and system integration
• TQ Analyst chemometric software for model building (SIMCA, PCA, discriminant analysis)
• Barcode reader for automated sample metadata capture; OPC interface for DCS communication

Main results and discussion

• Material identification: The Antaris II plus RESULT successfully discriminated many raw-material classes rapidly and non-destructively, even when spectra contained contributions from packaging liner polymers.
• Particle-size sensitivity: Spectral offsets due to scattering from different particle-size fractions were observed. The workflow conditionally invoked particle-size analysis only when the sample identity matched lactose in the library.
• Particle-size methods: Two classification approaches were developed: (1) Discriminant analysis on untreated spectra using SIMCA (requiring several standards per class) and (2) second derivative pretreatment with Norris smoothing to improve discrimination. Both approaches omitted polymer liner bands from the analysis.
• Sample set for particle size: Pharmatose lactose samples across mesh sizes 50 µ, 80 µ, 90 µ, 100 µ, 110 µ and 125 µ were used to demonstrate discrimination and detection of size-related spectral differences.
• Automation and integration: Barcode-based metadata capture, automated measure events, conditional Perform If logic and OPC messaging enabled hands-off operation with PASS/FAIL routing, archival, and plant-wide notification on failures.

Benefits and practical applications

• Speed and throughput: Seconds-scale FT-NIR identification accelerates goods-in checks compared with titration or HPLC and reduces hold times at the dock.
• Non-destructive testing: Ability to analyze through common packaging reduces waste and handling.
• Reduced operator skill requirement: Workflow automation and library-based classification enable use by warehouse personnel without spectroscopy expertise.
• Process integration: OPC and RESULT-driven decisions allow direct linkage to DCS/LIMS for automatic quarantine, release, and supplier notification workflows.
• Quality control expansion: Combining chemical class ID with conditional physical assessments (e.g., particle-size classification) enriches acceptance criteria beyond simple identity tests.

Future trends and potential uses

• Advanced chemometrics and machine learning: Use of supervised and deep-learning models could improve robustness to packaging variability and subtle polymorphic differences.
• Extended at-line and in-line deployment: Miniaturized FT-NIR probes and more robust sampling accessories can enable continuous monitoring at transfer points or inline process steps.
• Expanded libraries and shared models: Centralized, curated spectral libraries across suppliers and sites would support faster onboarding of new materials.
• Regulatory and data integrity focus: Tighter integration with electronic batch records, audit trails, and validated chemometric models will be required as PAT moves into regulated QC release roles.
• Hybrid measurements: Combining FT-NIR with complementary sensors (Raman, particle size analyzers) in unified workflows can provide richer descriptors of incoming materials.

Conclusions

The Antaris II FT-NIR Analyzer coupled with RESULT software provides an effective, automated solution for raw-material qualification at goods-in. The system demonstrated reliable chemical classification through packaging, detection of particle-size–related spectral effects, and conditional decision-making integrated with plant control and inventory systems. Such workflows reduce time-to-release, lower laboratory burden, and create a scalable PAT front end that supports quality-by-design principles in pharmaceutical and chemical manufacturing.

Reference

• Hirsch J. Raw materials qualification within a workflow: FT-NIR analysis using the Antaris II Analyzer. Thermo Fisher Scientific Application Note AN51088_E (2022).